Pancreatic ductal adenocarcinoma (PDA) is the fourth leading cause of cancer-related death worldwide, with a five-year survival rate of 4%. Because existing therapies have little impact on PDA patient survival, there is great urgency to identify new therapeutic targets for this lethal cancer. More than 90% of PDAs have oncogenic mutations in the Kras gene that initiate tumorigenesis and whose continued presence is required for PDA progression. Unfortunately, direct targeting of the Kras oncogene for cancer treatment has thus far not been successful. Phosphoinositide 3-kinase (PI3K) is a critical downstream effector of Kras involved in oncogenesis. We demonstrated that pancreas-specific ablation of the PI3K p110? catalytic subunit (officially named Pik3ca) completely protected mice from oncogenic KrasG12D-induced tumor formation. We then produced a novel mouse model to test whether Pik3ca is also required for pancreatic tumor maintenance. This mouse strain has a tetracycline-regulated Pik3ca gene that can be turned on or off in the pancreas of the KrasLSL-G12D/+ pancreatic tumor model. KrasLSL-G12D/+ mice develop the full spectrum of pancreatic intraepithelial neoplasias (PanINs) seen in human pancreatic cancer, but these early lesions infrequently progress to invasive PDA in this mouse model. Our preliminary data show that silencing of Pik3ca resulted in complete regression of early stage pancreatic tumors. However, as malignancies progress, accumulation of secondary mutations and epigenetic changes can result in resistance to a particular therapy. It is not known whether invasive or metastatic pancreatic cancer remains dependent on Pik3ca. Our driving hypothesis is that the dependence of Kras-induced pancreatic cancer on Pik3ca for continued growth changes during disease progression. To address our hypothesis, we will create a model of aggressive PDA by introducing the Trp53R172H mutation into mice expressing KrasG12D and inducible Pik3ca. The effect of silencing Pik3ca on three specific stages of pancreatic carcinogenesis will be investigated. Early Stage is defined as carcinoma in situ, i.e., up to PanIN grade 3, where the basement membrane is intact. Advanced Stage is defined as the appearance of gross pancreatic lesions or frank PDA with invasion but without metastatic lesions. Metastatic Stage is defined by the presence of metastatic disease, where we shall compare primary pancreatic tumors with metastatic deposits. The phenotypic features, transcriptomes and exome profiles of pancreatic tumors in this mouse model will be correlated with their dependence on Pi3kca in terms of tumor regression and survival time. Successful completion of this application will produce a useful animal model and molecular insight into the mechanisms of pancreatic tumor regression and potential resistance to Pik3ca inhibition. This understanding will allow future studies to elucidate the most appropriate means of targeting the Pik3ca signaling pathway and the oncogenic program regulated by this pathway for PDA therapy.